XPK Start: Fri Apr 17 13:12:22 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-17 13:12:46.871606: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)
I0417 13:12:47.054577 136716054284096 max_utils.py:273] Attempting to initialize the jax distributed system...
I0417 13:12:56.093798 136716054284096 distributed.py:149] Starting JAX distributed service on [::]:8482
I0417 13:12:56.096164 136716054284096 distributed.py:172] Connecting to JAX distributed service on mt-03-dropout-4wx7e-slice-job-0-0.mt-03-dropout-4wx7e:8482
I0417 13:12:57.240298 136716054284096 max_utils.py:284] Jax distributed system initialized!
I0417 13:13:03.541805 136716054284096 max_utils.py:800] System Information: Jax Version: 0.9.2
I0417 13:13:03.541925 136716054284096 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0417 13:13:03.541966 136716054284096 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0417 13:13:03.542001 136716054284096 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0417 13:13:04.219131 136716054284096 maxtext_utils.py:1548] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0417 13:13:04.219405 136716054284096 checkpointing.py:677] Setting up checkpoint logger...
I0417 13:13:04.219458 136716054284096 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0417 13:13:04.219501 136716054284096 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0417 13:13:04.219829 136716054284096 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7c571e7e29c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0417 13:13:14.446966 136716054284096 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0417 13:13:14.447211 136716054284096 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c430c150560>}, handler_registry=None
I0417 13:13:14.447484 136716054284096 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c430c150560>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`.
I0417 13:13:14.447539 136716054284096 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c430c1537d0>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`.
I0417 13:13:14.447576 136716054284096 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c430c150560>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c430c150560>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c430c1537d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c430c1537d0>}).
I0417 13:13:14.447971 136716054284096 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0417 13:13:14.448064 136716054284096 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c428437cae0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0417 13:13:18.197119 136716054284096 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints
I0417 13:13:18.836467 136716054284096 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c430c152c60>
I0417 13:13:18.836637 136716054284096 checkpointing.py:301] Checkpoint manager created!
I0417 13:13:19.164744 136716054284096 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 13:13:19.164876 136716054284096 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0417 13:13:19.543996 136716054284096 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 13:13:19.544086 136716054284096 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0417 13:13:19.560599 136716054284096 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 13:13:19.560655 136716054284096 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0417 13:13:19.584630 136716054284096 attentions.py:1154] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 13:13:19.584695 136716054284096 attentions.py:1154] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0417 13:13:19.601170 136716054284096 attentions.py:1155] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 13:13:19.601232 136716054284096 attentions.py:1155] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0417 13:13:19.617600 136716054284096 attentions.py:1156] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 13:13:19.617661 136716054284096 attentions.py:1156] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0417 13:13:19.643510 136716054284096 attentions.py:1197] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0417 13:13:19.643578 136716054284096 attentions.py:1197] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0417 13:13:19.669881 136716054284096 linears.py:525] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0417 13:13:19.669950 136716054284096 linears.py:525] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0417 13:13:19.896651 136716054284096 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0417 13:13:19.896751 136716054284096 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0417 13:13:21.329031 136716054284096 maxtext_utils.py:1651] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0417 13:13:21.329164 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0417 13:13:21.329220 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0417 13:13:21.329278 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0417 13:13:21.329332 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0417 13:13:21.329373 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0417 13:13:21.329428 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0417 13:13:21.329483 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0417 13:13:21.329526 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0417 13:13:21.329563 136716054284096 maxtext_utils.py:1651] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0417 13:13:21.329612 136716054284096 maxtext_utils.py:1651] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0417 13:13:21.329662 136716054284096 maxtext_utils.py:1651] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0417 13:13:21.853233 136716054284096 train.py:155] train/xent Logical: float32[32,128]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0417 13:13:21.853326 136716054284096 train.py:155] train/xent Physical: float32[32,128]............................................. ('fsdp', None).
I0417 13:13:21.868904 136716054284096 train.py:162] train/z_loss Logical: float32[32,128]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0417 13:13:21.868966 136716054284096 train.py:162] train/z_loss Physical: float32[32,128]............................................. ('fsdp', None).
I0417 13:13:25.379510 136716054284096 max_utils.py:791] Total memory size: 0.8 GB, Output size: 0.4 GB, Temp size: 0.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0417 13:13:25.387043 136716054284096 metric_logger.py:301] number parameters: 1.104 billion
I0417 13:13:29.593163 136716054284096 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0417 13:13:29.666707 136716054284096 checkpointing.py:776] Waited 0.07352900505065918 seconds for step 0 to finish before starting checkpointing.
I0417 13:13:29.669097 136716054284096 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 13:13:29.671295 136716054284096 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0
I0417 13:13:29.672695 136716054284096 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0.
I0417 13:13:30.953508 136716054284096 signaling_client.py:364] Using JaxDistributedSignalingClient
I0417 13:13:30.967745 136716054284096 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 13:13:30.967810 136716054284096 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0417 13:13:31.237944 136716054284096 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.270879s
I0417 13:13:31.238892 136716054284096 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.557 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2776069641113281 s) (per-host)
I0417 13:13:31.238968 136716054284096 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.277703s (batch_requests_ready=0.002369s, total_serialization_initiated=0.274474s, others=0.000860s)
I0417 13:13:31.239052 136716054284096 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.281658s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.281640)
I0417 13:13:31.239884 136716054284096 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.57 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0.
I0417 13:13:31.240167 136588407334656 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 13:33:31.240134
I0417 13:13:31.256740 136716054284096 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 13:13:31.256996 136587287443200 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0417 13:13:31.257125 136716054284096 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776431609.669079, 'wait_for_prev_duration_secs': 6.0558319091796875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776431609.6713336, 'checkpointer_blocking_duration_secs': 1.5689802169799805, 'get_old_steps_start_time': 1776431611.2403383, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776431609.6673083, 'checkpoint_manager_blocking_duration_secs': 1.5897784233093262}
I0417 13:13:31.257236 136716054284096 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0417 13:13:31.257287 136716054284096 max_utils.py:750]
Memstats: After params initialized:
I0417 13:13:31.257336 136716054284096 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_0(process=0,(0,0,0,0))
I0417 13:13:31.257370 136716054284096 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_1(process=0,(1,0,0,0))
I0417 13:13:31.257396 136716054284096 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_4(process=0,(0,1,0,0))
I0417 13:13:31.257420 136716054284096 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_5(process=0,(1,1,0,0))
I0417 13:13:31.588886 136716054284096 metric_logger.py:196] completed step: 0, seconds: 4.009, TFLOP/s/device: 0.200, Tokens/s/device: 31.931, total_weights: 4096, loss: 10.889, lm_loss: 10.889, perplexity: 53560.699
I0417 13:13:31.592215 136716054284096 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/tensorboard/'
I0417 13:13:32.367413 136588398941952 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0
I0417 13:13:32.475146 136716054284096 metric_logger.py:196] completed step: 1, seconds: 1.994, TFLOP/s/device: 0.402, Tokens/s/device: 64.197, total_weights: 4096, loss: 10.902, lm_loss: 10.902, perplexity: 54298.523
I0417 13:13:32.543889 136716054284096 metric_logger.py:196] completed step: 2, seconds: 0.887, TFLOP/s/device: 0.903, Tokens/s/device: 144.319, total_weights: 4096, loss: 9.956, lm_loss: 9.956, perplexity: 21082.051
I0417 13:13:32.614485 136716054284096 metric_logger.py:196] completed step: 3, seconds: 0.010, TFLOP/s/device: 82.829, Tokens/s/device: 13238.184, total_weights: 4096, loss: 9.139, lm_loss: 9.139, perplexity: 9314.691
I0417 13:13:39.039979 136588382156544 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/items
I0417 13:13:39.637324 136586249910016 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776431618270287791, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776431618270287791, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/_CHECKPOINT_METADATA
I0417 13:13:43.408053 2808 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0417 13:13:48.975339 136587304228608 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/items/array_metadatas/process_0
I0417 13:13:49.794901 136716054284096 metric_logger.py:196] completed step: 4, seconds: 0.069, TFLOP/s/device: 11.569, Tokens/s/device: 1848.936, total_weights: 4096, loss: 8.459, lm_loss: 8.459, perplexity: 4715.493
I0417 13:13:49.803976 136716054284096 metric_logger.py:196] completed step: 5, seconds: 0.071, TFLOP/s/device: 11.320, Tokens/s/device: 1809.187, total_weights: 4096, loss: 7.924, lm_loss: 7.924, perplexity: 2762.373
I0417 13:13:49.872766 136716054284096 metric_logger.py:196] completed step: 6, seconds: 17.179, TFLOP/s/device: 0.047, Tokens/s/device: 7.451, total_weights: 4096, loss: 7.553, lm_loss: 7.553, perplexity: 1906.617
I0417 13:13:49.943422 136716054284096 metric_logger.py:196] completed step: 7, seconds: 0.007, TFLOP/s/device: 118.966, Tokens/s/device: 19013.666, total_weights: 4096, loss: 7.267, lm_loss: 7.267, perplexity: 1431.653
I0417 13:13:50.014061 136716054284096 metric_logger.py:196] completed step: 8, seconds: 0.070, TFLOP/s/device: 11.521, Tokens/s/device: 1841.382, total_weights: 4096, loss: 7.111, lm_loss: 7.111, perplexity: 1225.983
I0417 13:13:50.084167 136716054284096 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0417 13:13:50.084828 136716054284096 checkpointing.py:776] Waited 0.0006783008575439453 seconds for step 9 to finish before starting checkpointing.
I0417 13:13:50.086827 136716054284096 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 13:14:20.423963 136587295835904 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.018840s (commit=33.946811s, array_metadata_write=3.072029s)
I0417 13:14:20.426041 136588407334656 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 31.934 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 49.46472883224487 s) (per-host)
I0417 13:14:20.426176 136588407334656 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 49.185896s.
I0417 13:14:22.509952 136588407334656 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776431618270287791, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/_CHECKPOINT_METADATA
I0417 13:14:24.779307 136588407334656 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0417 13:14:24.963159 136586249910016 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776431618270287791, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/_CHECKPOINT_METADATA
I0417 13:14:25.370111 136588407334656 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000184s.
I0417 13:14:25.473111 136588407334656 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.406736s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/items
I0417 13:14:25.474725 136588407334656 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0/items
I0417 13:14:26.562407 136588407334656 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0
I0417 13:14:31.617470 136588407334656 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0`.
I0417 13:14:31.618237 136588407334656 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 61.95 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/0
I0417 13:14:31.619453 136588407334656 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 60.379156s.
I0417 13:14:31.619683 136587287443200 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 13:14:31.619765 136587287443200 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0417 13:14:31.619832 136587287443200 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0417 13:14:31.622795 136587287443200 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0417 13:14:31.623001 136716054284096 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0417 13:14:31.623147 136716054284096 checkpoint_manager.py:1452] Waiting for previous save to complete took 41.536311 seconds. If this number is high, consider checkpointing less frequently.
I0417 13:14:31.625068 136716054284096 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9
I0417 13:14:31.627004 136716054284096 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9.
I0417 13:14:32.776156 136716054284096 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 13:14:32.776248 136716054284096 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0417 13:14:32.811483 136716054284096 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.036076s
I0417 13:14:32.812270 136716054284096 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 37.865 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04073929786682129 s) (per-host)
I0417 13:14:32.812339 136716054284096 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.040825s (batch_requests_ready=0.001969s, total_serialization_initiated=0.038146s, others=0.000709s)
I0417 13:14:32.812418 136716054284096 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.044752s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.044734)
I0417 13:14:32.813139 136716054284096 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.19 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9.
I0417 13:14:32.813437 136587287443200 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 13:34:32.813404
I0417 13:14:32.815320 136716054284096 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 13:14:32.815584 136587845285632 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0417 13:14:32.815721 136716054284096 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776431630.086795, 'wait_for_prev_duration_secs': 41.53631114959717, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776431671.625108, 'checkpointer_blocking_duration_secs': 1.188462734222412, 'get_old_steps_start_time': 1776431672.8135946, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776431630.0850701, 'checkpoint_manager_blocking_duration_secs': 42.73061776161194}
I0417 13:14:32.815901 136716054284096 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0417 13:14:32.815950 136716054284096 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 13:14:33.436876 136588407334656 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9
I0417 13:14:38.467584 136587853678336 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9/items
I0417 13:14:45.189889 136587870463744 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9/items/array_metadatas/process_0
I0417 13:15:23.946059 136587862071040 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.962510s (commit=38.412002s, array_metadata_write=3.550508s)
I0417 13:15:23.947132 136587287443200 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 30.866 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 51.17557096481323 s) (per-host)
I0417 13:15:23.947257 136587287443200 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 51.133711s.
I0417 13:15:27.694340 136587287443200 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0417 13:15:28.053019 136587287443200 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000256s.
I0417 13:15:28.618736 136587287443200 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.400118s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9/items
I0417 13:15:28.620564 136587287443200 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9/items
I0417 13:15:29.601065 136587287443200 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9
I0417 13:15:32.289372 136587287443200 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9`.
I0417 13:15:32.290202 136587287443200 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 60.67 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_03_dropout/checkpoints/9
I0417 13:15:32.291746 136587287443200 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 59.478197s.
I0417 13:15:32.291911 136587845285632 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 13:15:32.291968 136587845285632 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0417 13:15:32.292027 136587845285632 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0417 13:15:32.294441 136587845285632 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0417 13:15:32.295507 136716054284096 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0417 13:15:32.295663 136716054284096 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 13:15:32.296669 136716054284096 metric_logger.py:196] completed step: 9, seconds: 0.071, TFLOP/s/device: 11.339, Tokens/s/device: 1812.312, total_weights: 4096, loss: 6.981, lm_loss: 6.981, perplexity: 1075.867
Per train step:
Total TFLOPs: 0.80
split as 99.60% learnable weight flops and 0.40% attention flops
XPK End: Fri Apr 17 13:15:43 UTC 2026
EXIT_CODE=0